Sources and method

What is calculated

Midsummer sunlight at 65°N

The number is the daily mean of incoming solar radiation at 65°N on the northern summer solstice, at the top of the atmosphere. It is not sunlight at the ground, a temperature, or an ice-sheet size. The lab’s temperature and ice estimates are a separate, rougher step, described below.

The calculation uses a total solar irradiance of 1361 W/m² and the standard daily insolation geometry from Berger's astronomical formulation.

The clock

800,000 years of orbit and ice

The orbit on the clock is the nominal La2004 solution of Laskar and colleagues, read at every 1,000 years from 800,000 years ago to 100,000 years ahead, with linear steps in between. The four moments above are exact nodes of the same data.

The ice line is the LR04 stack of 57 benthic oxygen-isotope records (Lisiecki and Raymo, 2005), which follows global ice volume. The scale runs from 0% at today’s value to 100% at its highest point in the last 30,000 years. Beyond 600,000 years the stack is sampled every 2,000 years and the gaps are filled in linearly. The line stops at today because nothing here predicts future ice.

The ice drawn on the globe is an illustration. Its size follows the ice line, and its outline is drawn by hand after the last glacial maximum. Earth’s surface comes from NASA Earth Observatory imagery: Blue Marble, Black Marble and GEBCO terrain.

Temperature and ice

A rough estimate, in two parts

At a date on the clock, ice is the measured LR04 value. Ice and global temperature move together, and the last glacial maximum was about 6.1 °C colder than before industry (Tierney et al., 2020), so temperature is minus 6.1 °C times the ice share. A figure near 14 °C is used for the average temperature before industry.

For an orbit set by hand, the lab asks what would settle if it lasted. Fitted to the last 800,000 years, ice follows 65°N summer sunlight smoothed over about 15,000 years, at 0.018 of the ice-age peak for every W/m² of sunlight lost (correlation about −0.46). On the warm side the response flattens, because there is little ice left to melt. The fit is checked against the data in the project’s tests.

Carbon dioxide, oceans and the ice already in place are left out, so the estimate can differ a lot from a real date: with a hand-set orbit, 21,000 years ago shows far less ice than the record, because the record built up over many earlier summers. Today’s real temperature is about 1.2 °C above the baseline because of greenhouse gases.

Limits

Sunlight alone does not predict climate

The lab calls sunlight within ±5 W/m² of the J2000 reference “close to today’s level”. That range is a wording choice, not a physical threshold.

Existing ice, snowfall, oceans, greenhouse gases, vegetation, dust, geography and long response times decide what the climate does. The +50,000-year moment is orbital geometry, not a climate forecast.